Dongchedi scraping collects listings and vehicle content from China’s major auto platform (懂车帝, ByteDance), where prices are in CNY, discovery is app-first, and B2B access is available via the open partner platform op.dongchedi.com. VivaCoding designs compliant pipelines: mobile/web endpoints, Chinese trim normalization, dedup, schedules, and CRM or catalog export — prioritizing official partner APIs where contracts exist, not gray captcha bypass.
Dongchedi is ByteDance’s auto vertical blending media, classifieds, and tooling for new/used cars in China. It aggregates dealers, private listings, video reviews, and configurators; much of the UX lives in the mobile app. Exporters, analytics teams, and international dealers of Chinese brands (BYD, Geely, Chery, NIO, etc.) need structured price and trim monitoring — not app screenshots. Below: ecosystem overview, fields, app-first technical challenges, and custom scraper vs partner API comparison.
Key takeaways
- Dongchedi is ByteDance’s auto vertical; China market; CNY; app-first discovery.
- Open platform op.dongchedi.com serves partners (dealers/media), not a free whole-market public API.
- Data spans new/used cars, specs, sometimes video metadata — CN→global taxonomy normalization required.
- App APIs use anti-bot and signed requests — pipeline needs partner keys or legally scoped custom fetch.
- VivaCoding: partner API first, custom fetch second, ETL + data collection to your stack.
Dongchedi platform overview
Dongchedi competes with Autohome and Yiche for auto content and leads. ByteDance strength: recommendation feed, short video, Douyin tie-in. For data buyers, value is in current dealer quotes, regional pricing (tier-1 vs tier-3 cities), and rich Chinese EV specs (battery, CLTC range).
op.dongchedi.com is the open platform for certified partners: dealer systems, media integrations, sometimes inventory sync. With an existing partner account we start on documented endpoints; without one we scope custom collection and legal basis — we do not promise a “full China dump.”
Web dongchedi.com exists, but many listing details and price updates appear first in app JSON APIs — we design mobile-aware pipelines with device management only under an agreed model.
China’s 2024–2026 auto market is a price war among BYD, Tesla CN, and legacy JVs. Export buyers track dealer discount vs MSRP in tier-1 cities (Shanghai, Shenzhen) vs inland — spreads of 8–15% on some EVs. Dongchedi aggregates dealer quotes faster than static OEM MSRP pages. ByteDance personalization shows different inventory to different users; pipelines rely on search/list APIs, not personalized feeds.
For European importers we store CLTC range separately from WLTP-equivalent estimates with disclaimers — never replacing official EU homologation. Battery kWh, charging standard (GB/T vs CCS), and RHD availability are separate export columns for customs and marketing.
Data fields we collect
| Field | Use case | Notes |
|---|---|---|
| vehicle_id / sku_id | Dedup | App vs web ID mapping |
| brand, series, model (CN) | Catalog | BYD 汉 → Han EV mapping |
| price_cny, msrp, discount | Margin analytics | Dealer quote vs MSRP |
| city, province | Regional arbitrage | Tier city classification |
| fuel_type, ev_range_cltc | EV export specs | CLTC vs WLTP note |
| registration_year, mileage | Used segment | km, plate city codes |
| photos[], video_id | Merchandising | CDN; optional skip video bytes |
| dealer_id, store_name | Lead routing | PII policy |
| config_options[] | Trim matrix | CN packs → EN labels |
| updated_at, status | Freshness | Off-sale detection |
Technical challenges
App-first APIs. Listing JSON often needs headers (device id, app version, signatures). Without partner keys the layer is fragile — we version client profiles and monitor 4xx spikes.
Chinese taxonomy. Series/model names differ from export naming — maintained mapping table required.
CLTC vs global standards. Store raw CN labels plus normalized fields with disclaimers for EU/US catalogs.
Latency. Fetch from EU infra is slow; optional CN egress or partner CDN — decided at brief.
Regulatory. Cross-border auto data may trigger PIPL — storage and transfer assessment for EU clients.
Pipeline architecture
- Access model — partner API keys from op.dongchedi.com OR scoped public fetch.
- Legal & PIPL — purpose limitation; skip personal plates if unnecessary.
- Discovery — city × model matrix; separate used vs new channels.
- Fetch — app API adapter + web fallback; signed request rotation.
- Normalize — CNY, CN trim dictionary, CLTC metadata.
- Store & diff — price history, delist rules.
- Export — to automotive site, BI, SFTP.
Custom scraper vs partner API
| Criteria | Partner API (op.dongchedi.com) | Custom fetch (VivaCoding) |
|---|---|---|
| Onboarding | Needs Chinese entity/partner status | Faster start for read-only analytics |
| Stability | Documented, supported | Depends on app releases |
| Coverage | Usually dealer-bound inventory | Broader search slices if legal |
| Cost | Partner fees | Dev + maintenance |
| Compliance | Contractual | Explicit legal review |
| EV spec depth | Often complete for own stock | Full market if scoped |
Best fit is often hybrid: partner feed for owned dealers + custom benchmark monitor. We compare commercial “China car APIs” fairly — many lag 24–48h with incomplete used segments.
When this service fits
- Chinese EV/PHEV importer — monitor dealer quotes by model and city.
- Media/research — BYD/Tesla CN price index vs export.
- Dealer network with op.dongchedi.com access — sync to global catalog.
- Multi-market scraping hub (CN + KR Encar + EU AutoScout24).
When it does not fit
- Full Dongchedi mirror without legal/partner path — we decline.
- Expecting a stable free public whole-market API — not available.
- One-time pull of 5 models — manual export suffices.
- Bypassing real-name auth or paywalls — not our scope.
Timeline and budget estimates
| Scope | Timeline | Budget from |
|---|---|---|
| MVP: web channel, 10 models, CSV | 3–4 weeks | €2,190 |
| App API adapter, CN dictionary, dedup | 6–8 weeks | €4,490 |
| Partner API + multi-city SLA | 10–14 weeks | €8,490+ |
CN egress, dictionary updates — from €390/mo. See web scraping.
Related services
- web scraping
- automotive websites
- data collection
- catalog website development
- website development
- contact
FAQ
What is op.dongchedi.com?
Dongchedi’s open partner platform: dealer tools, integrations, contracted API access — not a public whole-market Google-style API.
Can we scrape web only, without the app?
Partially for new-car specs. Used segment and fresh quotes are often richer in the app — we recommend hybrid.
How are Chinese model names translated?
Maintained mapping table + OEM export names; optional human QA for marketing copy.
Dongchedi vs Autohome?
Autohome is the legacy leader; Dongchedi has ByteDance growth and video. Separate adapters, shared ETL.
Is collection from China to EU legal?
PIPL + GDPR assessment; minimize personal data (plates, phones). Document lawful basis before start.
CRM integration?
Webhook/REST to Bitrix24, AmoCRM, or custom via our ETL layer.
Refresh frequency?
Dealer quote monitoring: every 4–12 hours. Research index: daily. Real-time only via official partner stream if available.